Innate immune recognition of an AT-rich stem-loop DNA motif in the Plasmodium falciparum genome

Abstract

Although Toll-like receptor 9 (TLR9) has been implicated in cytokine and type I interferon (IFN) production during malaria in humans and mice, the high AT content of the Plasmodium falciparum genome prompted us to examine the possibility that malarial DNA triggered TLR9-independent pathways. Over 6000 ATTTTTAC ("AT-rich") motifs are present in the genome of P. falciparum, which we show here potently induce type I IFNs. Parasite DNA, parasitized erythrocytes and oligonucleotides containing the AT-rich motif induce type I IFNs via a pathway that did not involve the previously described sensors TLR9, DAI, RNA polymerase-III or IFI16/p204. Rather, AT-rich DNA sensing involved an unknown receptor that coupled to the STING, TBK1 and IRF3-IRF7 signaling pathway. Mice lacking IRF3, IRF7, the kinase TBK1 or the type I IFN receptor were resistant to otherwise lethal cerebral malaria. Collectively, these observations implicate AT-rich DNA sensing via STING, TBK1 and IRF3-IRF7 in P. falciparum malaria.

Figure 1. P. falciparum induced interferon stimulated gene expression in human cells

A. A microarray analysis of PBMCs from 14 patients infected with Pf, before or after curative treatment with mefloquine. A selection of ISGs from a cohort of 580 upregulated genes is depicted. B. PBMCs were stimulated with Pf-infected red blood cells (iRBC), uninfected RBCs (RBCs) or SeV (200HAU/ml) for 3 to 9 hours. C. PBMCs were stimulated with LPS (100ng/ml), 100μM natural hemozoin (Hz) untreated or pretreated with S. aureus micrococcal nuclease (Hz + DNAse) or 5 μg of genomic DNA (Pf gDNA) complexed with lipofectamine 2000. IFNβ and β-actin mRNA was measured by qPCR. D. Mouse BMDM were treated with 200μM Hz and Hz and lysosomes were visualized with acousto-reflection microscopy and lysotracker respectively. Fields are representative of at least 10 fields of view.

Figure 2. Immune stimulatory activity of AT-rich motifs from P. falciparum

A. IFNβ reporter gene activity was measured from cells stimulated with media alone (−), purified Pf g DNA (5μg/ml), AT5 rich ODN (20-mer, 3μM, phosphorothiorate backbone (PS)), Pf chromosomal-derived Pf Chr 9_2 long ODN (56-mer, 3μM, PS), poly (dA-dT) (5μg/ml) complexed with lipofectamine 2000. Data are represented as fold induction relative to the reporter-only control and reflect mean fold induction ± SD. B. BMDM were stimulated with AT-rich ODN on phosphodiester backbones (PD), SeV and pdAdT. IFNβ secretion was measured 16h later by ELISA. C–I. HEK293, PBMCs, THP-1, mouse splenocytes, embryonic fibroblasts (MEFs), BMDMs and splenic dendritic cells (panDCs) were stimulated for 6h with 3μM AT2/AT5 (PS) or AT5 (PD) and IFNβ levels measured by qPCR. J. NF-κB reporter gene activity in HEK293 cells stimulated with AT2 (3μM, PS), poly (dA-dT) (5μg/ml), or infected with SeV (200HAU/ml). K. IL-6 or TNFα levels were measured 18h later by ELISA from macrophages treated as indicated. Data are presented as mean ± SD and are representative of 3 independent experiments.

Figure 3. AT-rich motifs induce IFNβ production in a sequence and secondary structure-dependent manner

A. HEK293s transfected with the IFNβ reporter plasmid were stimulated with SeV, poly(dA-dT) and AT-rich ODN that were modified (*) as in Sup Fig. 2a and luciferase activity was read an additional 24h later. B. Putative secondary structure of AT-rich ODN AT5, with AT-motif and likely chief base-base interactions highlighted. Key base-base interaction and secondary structure was abolished by base-directed substitution of the original sequences as shown in Sup Fig. 2b. Splenocytes from C57BL/6 mice were stimulated with the parent AT5 (3μM, PD) and denoted base substituted oligonucleotides and IFNβ mRNA induction measured by qPCR. C. Predicted secondary structure of annealed AT-5 and reverse strand and ISD (3μM each, PD). BMDMs from C57BL/6 mice were stimulated with the ssAT5, AT5 annealed to its reverse strand (AT5 + R), annealed ISD and ssISD (forward strand only) for 18h and IFNβ secretion was measured by ELISA.

Figure 4. AT-rich DNA mediated IFNβ induction does not require TLRs, cytosolic sensors or the adaptor protein MAVS

A–B. Splenocytes from the indicated strains were stimulated with CpG-B 1826 (5μM), AT5 (3μM, PD) or with Pf 3D7 genomic DNA (100ng/ml) complexed with lipofectamine 2000 for 6h and IFNβ measured by qPCR and/or ELISA C–G. BMDMs from the indicated strains were stimulated with lipofectamine complexed AT5 (3μM, PD), poly (dA-dT) (5μg/ml), poly (I:C) (100ng/ml) or infected with SeV (200HAU/ml) and IFNβ mRNA measured by qPCR or ELISA. H. HEK293 cells were transduced with 2 lentiviruses encoding shRNA for DAI and IFNβ reporter gene activity measured measured from cells stimulated with lipofectamine complexed-AT2 (3μM, PS) or infected with NDV (80HAU/ml) for 24h. The efficiency of DAI silencing was examined by RT-PCR.

Figure 5. P. falciparum AT-rich motifs do not signal via RNA-polymerase III/RIG-I

A. RIG-I, MAVS and RNA polymerase III component 1 (RPC1) were silenced by RNAi. 48h later, cells were stimulated with AT2 (PS; left side of graphic), poly (dA-dT) and SeV (right side of graphic). IFNβ gene induction was assessed 24h later with qPCR B. HEK293 cells were transfected with either poly (dA-dT) or AT2 and RNA from the transfected cells was harvested 16 hr later. Harvested RNA was fractionated and purified as indicated and transfected back into PBMCs and IFNα production assessed 9 or 18h later. C. HEK293 cells were pre-treated with 0, 2.5, 5, 10, 20 or 30ng of the RNA polymerase III inhibitor, ML-60218 or vehicle alone (DMSO) for 2h and IFNβ reporter activity measured after AT2 (PS) or poly (dA-dT) treatment. D. 0, 20, 40, 60 and 80ng of plasmids encoding the Hepatitis C Virus (HCV) protease NS3/4A and the inactive mutant NS3/4A S139A were transfected into HEK293 cells along with the IFNβ reporter. 24h later the cells were stimulated with AT2 (PS), poly (dA-dT) and SeV. Luciferase activity was assessed a further 24 h later. E. C57Bl6 or DAI^−/− macrophages were pretreated with ML-60218 for 2h and then stimulated as indicated and IFNβ levels measured by ELISA. F. DAI was silenced by RNAi and cells pretreated with the ML60218 inhibitor. IFNβ reporter gene activity measured after cells were stimulated with AT2, poly(dA-dT) and Sendai virus (SEV). shRNA mediated knockdown of DAI (ZBP-1) was assessed using western blotting at 48hrs after transfection of shRNAs. G. Mouse macrophages were electroporated with siRNA to Ifi204 or scrambled siRNA control for 48hrs and then stimulated with indicated ligands for a further 24hrs. Supernatants were assessed for levels of IFNβ (top) by ELISA, while lysates of cells were assessed for levels of Ifi204 remaining (bottom) by western blotting.

Figure 6. AT-rich DNA mediated IFNβ induction uses the adaptor STING, TBK1 and the interferon regulatory factors IRF3 and IRF7

A. BMDM from Tmem173^−/− mice were transfected with lipofectamine complexed AT5 (3μM, PD), Pf 3D7 g DNA (Pf gDNA,100ng/ml), poly (dA-dT) (3.5μg/ml) or infected with SeV (200HAU/ml) and IFNβ gene induction measured 6h later by qPCR. B. STING was silenced by siRNA and IFNβ reporter gene activity measured after treatment with poly (dA-dT) and AT2. C. Bone marrow derived macrophages from Tmem173^+/+ and Tmem173^−/− mice were stimulated as indicated for 6hrs and gene expression analyzed using Nano-string technology. D. Tbk1^−/− Tnfr1^−/− and Tnfr1^−/− BMDM were treated as indicated and IFNβ mRNA levels measured by qPCR. E. HEK293 cells (upper panels) or BMDMs from Tnfr1^−/−, Tbk1^−/− Tnfr1^−/−, Tmem173^+/+ and Tmem173^−/− mice were transfected as indicated with biotinylated ODNs. Cell lysates were subjected to pull-down analysis and immunoblots were probed for the presence of Sting and TBK1 by western blotting. F. HEK293 cells were transfected with multimerized PRDIII-I and PRDII reporter elements from the IFNβ promoter and stimulated as above. Luciferase reporter activity was measured. G–H. BMDMs from Irf3^−/− Irf7^−/−, Irf3^−/− and Irf7^−/− mice were transfected as indicated and levels of IFNβ mRNA were quantified 6h later by qPCR.

Figure 7. Role of STING, TBK1 and the transcription factors IRF3-IRF7 in Plasmodium induced IFN responses in vitro and in vivo

A. Pf infected RBCs (iRBCs) or uninfected RBCs (uRBCs) were incubated with C57Bl/6, Irf3^−/−-Irf7^−/−, Tbk1^−/−-Tnfr1^−/− and Tnfr1^−/− BMDMs as indicated (B–D). Plasmodium berghei ANKA (PbA) iRBCs were incubated with BMDMs from Irf3^−/−-Irf7^−/−, Tbk1^−/−-Tnfr1^−/−, Tnfr1^−/− and Tmem173^−/− mice and IFNβ induction measured by ELISA. (E–G). C57BL/6 (n=15), Ifnar^−/− (n=10),, Irf3^−/−/If7^−/− (n=7), Irf3^+/+irf7^+/+ (n=8), Tbk1^+/+ (n=5) and Tbk1Δ/Δ (n=5) mice were infected i.p. with 10⁵ parasites of P. berghei ANKA strain and mice monitored daily for survival as shown. (*** p ≤ 0.0001, student’s t-test). A representative of three independent experiments is shown.